I have notes on an article by a "Tukey" who introduces the term wicked
problem, I believe and defines it as a "Problem of the Third Kind." He
uses the concept from statistics of type 1 and type 2 errors to contrast
with type 3 errors in wicked problems: solving the wrong problem. I
believe that the article is titled errors of the third type or the third
kind. He lists characteristics of wicked problems. The following are MY
notes on these criteria. Here is what is missing in wicked problems:
1. ability to formulate the problem
2. relationship between problem and the solution
a) will it cause new problems, intertwined
3. testability, there is none
4. finality, solve aspects, but there is always more
5. tractability
a) no list of ways to solve it
6. explanatory characteristics
a) solutions take on different forms depend on how explain
problem, more than one explanation
7. Level of analysis
a) every problem is a symptom of another problem
8. reproducibility
a) no trial and error approximation
b) not conducive to modeling
9. replicability, each is unique
10. responsibility, problem solver is responsible for errors,
never gets praise
11. additions
a) leadership, basically leader is to hold anxiety, to be moral, create
appropriate fiction
(1) follower is reciprocal
b) Culture-diagnostic
(1) disparate assumptions are allowed, tolerated vs. unidimensional,
one right way
(2) diagnose if culture is tame or complex by look at how do problem solving
I hope that this is of enough interest that the bad ASCII translation and
length is tolerable.
By the way, another interesting article in the same vein is Lindblom
(1952), "The Science of 'Muddling Through'", Public Admin. Review. This
is about how policy makers "muddle through" rather than rationally solve
(wicked) policy decisions.
-- Lew Mills, MA mills@itsa.ucsf.edu